Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:121283
MIME-Version: 1.0
References: <CAEZPtU6PT7yOhVZfGQ1Vh3soKH0tkXh86v0GiNJo79ZMiuO4XQ@mail.gmail.com>
 <9812BAD1-5003-4E91-BB3B-00637B9687AC@sakiot.com> <CAMrTa2H553O1CesPU95ma-HYnejWkMgZf-QOVvCxs=NHMOAc+g@mail.gmail.com>
 <CAEZPtU4v8Z0=pX7+YSXYPp+TDUcdWBk-WYHkbySabMnyFGqHiQ@mail.gmail.com> <CAMrTa2GT9BD1=8hf5U98ZeTqot=qRMffHmkw30ww5GvELLQzbw@mail.gmail.com>
In-Reply-To: <CAMrTa2GT9BD1=8hf5U98ZeTqot=qRMffHmkw30ww5GvELLQzbw@mail.gmail.com>
Date: Fri, 13 Oct 2023 02:47:16 +0700
Message-ID: <CAEZPtU4y-YvUpsZPVSWjMmttCzAozxGbyMmyF+XaJw1UfU8Wmg@mail.gmail.com>
To: Jordan LeDoux <jordan.ledoux@gmail.com>
Cc: Saki Takamachi <saki@sakiot.com>, PHP internals <internals@lists.php.net>
Content-Type: multipart/alternative; boundary="000000000000b83bde06078a3537"
Subject: Re: [PHP-DEV] [RFC] [Discussion] Adding bcround, bcfloor and bcceil
 to BCMath
From: pierre.php@gmail.com (Pierre Joye)

--000000000000b83bde06078a3537
Content-Type: text/plain; charset="UTF-8"

On Fri, Oct 13, 2023, 2:01 AM Jordan LeDoux <jordan.ledoux@gmail.com> wrote:

>
>
>
> I believe MPFR is what is used by ext-decimal. The only thing that makes
> ext-decimal a somewhat difficult replacement for BCMath is that one of them
> uses precision and the other uses scale. (Total accurate digits versus
> total significant digits.) Fermat has implementations for both BCMath and
> ext-decimal, while also using GMP. It is much more difficult to control the
> number of accurate digits in ext-decimal than in BCMath, but ext-decimal is
> literally hundreds of times faster. I think, however, that Rudi (the author
> of ext-decimal) was against including ext-decimal in core.
>

it isn't really needed if not at the language level.

scaled basic operations can be (relatively) easily optimized, with an order
of magnitude faster than what bcmath does.

however, the real challenge, imho, is the different uses. That's where rock
solid implementation like mpfr could help, for non basic ops or modes.



> But if you want an alternate implementation for arbitrary precision
> decimals, it is certainly the place that someone should start.
>

for the few areas I needed it, I made my own with add/sub and mul/div with
intrinsics.  very little amount of codes and stable.

I am not completely sure how to cover the different usages, scaled
operations may be a good start and keeping bc math user land compatibility
may not too hard.

--000000000000b83bde06078a3537--